Quote from: Lutpin on July 01, 2019, 04:25:40 PM

Quote from: Timetwister on July 01, 2019, 07:59:33 AM

I used the second way of calculating it. These are the exact numbers (in Bitcoins):
July 25 2018. Bankroll: 1521.37. Profit: 1133.86
Today (now). Bankroll: 1281.73. Profit: 1276.78
So, assuming bankroll didn't change, investors should had made in that period of time: (1276.78-1133.86)/1521.37=9.39%
As bankroll kept dropping, profits for someone that stayed invested all the time would actually be higher.

Not necessarily. As the bankroll decreases in size, your share of it goes up, which means you share more of the profit, but equally share more of the losses of the house as well.
Lately, SuperStar has been winning a lot (a combined profit over 10BTC at the time of this writing), which are losses that are proportionally stemmed by all investments.

Let me explain this with an example:
Say you have 15BTC invested as part of a 1500BTC bankroll. For a fixed period of time, the profit for investors is 100BTC. You share 1% of that, increasing your investment to 16BTC.
Now the bankroll drops to 1000BTC. In the following period, investors make a loss of 20BTC. The chart would show 100BTC-20BTC= 80BTC of investors profit over both periods.
Your 16BTC investment shares 1.6% of that (0.32BTC), decreasing your investment size after both periods to 15.68BTC.
While the calculations over a fixed bankroll tell you to expect (100-20)/1500= 5.3%, in reality you see 15.68/15= 4.2%

This effect of losses at a later time compared to profit at an early time is increased by the fact that any profit made by your investment gets automatically added to it and becomes part of your investment.
That means that not only does the total bankroll decreases, giving you a bigger share of losses to cover, but also your individual investment increases, further driving up your share.

Side note: I thought I had already been pushing my accuracy with a 3 month period. 1 year is just too long a time to make a rough estimate over, imho.
If you want anything accurate, split the year into 4 or even better 6 or 12 parts and calculate expectations for each one of them.

I understand that mechanism. However, as during that period of time what happened is that the casino made money, instead of losing it, while bankroll decreased, profits are very likely to be higher than with a constant bankroll, and therefore higher than 9.39%. That's quite far from the 7.16% I made, and by coincidence, it's near 70% of the graph's profit if we consider they were higher than 9.39% as bankroll decreased (not constantly, but in general).

Anyway, I'll split that time in different periods as you say and see what's going on, but I suspect the graph is showing profits before fees.